Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

Dataset: 80% C4 (Web Text) and 20% Python Code

Neuron 2380 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 3.9152. Min Range: -3.9152

Max Act: 3.9152. Min Act: -0.0000

Data Index: 204963 (C4 (Web Text))

Max Activating Token Index: 97

Text #1

Max Range: 3.9152. Min Range: -3.9152

Max Act: 3.7835. Min Act: -0.0000

Data Index: 683542 (C4 (Web Text))

Max Activating Token Index: 284

Text #2

Max Range: 3.9152. Min Range: -3.9152

Max Act: 3.0193. Min Act: -0.0000

Data Index: 740363 (C4 (Web Text))

Max Activating Token Index: 83

Text #3

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.7909. Min Act: -0.0000

Data Index: 286278 (C4 (Web Text))

Max Activating Token Index: 12

Text #4

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.8265. Min Act: -0.0000

Data Index: 1441978 (Python Code)

Max Activating Token Index: 553

Text #5

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.9481. Min Act: -0.0000

Data Index: 1263283 (C4 (Web Text))

Max Activating Token Index: 51

Text #6

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.5325. Min Act: -0.0000

Data Index: 806856 (C4 (Web Text))

Max Activating Token Index: 548

Text #7

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.5632. Min Act: -0.0000

Data Index: 1256803 (C4 (Web Text))

Max Activating Token Index: 712

Text #8

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.2393. Min Act: -0.0000

Data Index: 694777 (C4 (Web Text))

Max Activating Token Index: 396

Text #9

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.2673. Min Act: -0.0000

Data Index: 1161184 (C4 (Web Text))

Max Activating Token Index: 322

Text #10

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.4362. Min Act: -0.0000

Data Index: 434757 (C4 (Web Text))

Max Activating Token Index: 117

Text #11

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.4392. Min Act: -0.0000

Data Index: 1036390 (C4 (Web Text))

Max Activating Token Index: 142

Text #12

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.2717. Min Act: -0.0000

Data Index: 1045781 (C4 (Web Text))

Max Activating Token Index: 274

Text #13

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.2558. Min Act: -0.0000

Data Index: 1486226 (Python Code)

Max Activating Token Index: 551

Text #14

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.3145. Min Act: -0.0000

Data Index: 318032 (C4 (Web Text))

Max Activating Token Index: 651

Text #15

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.1909. Min Act: -0.0000

Data Index: 292001 (C4 (Web Text))

Max Activating Token Index: 409

Text #16

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.2640. Min Act: -0.0000

Data Index: 899264 (C4 (Web Text))

Max Activating Token Index: 98

Text #17

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.1504. Min Act: -0.0000

Data Index: 268383 (C4 (Web Text))

Max Activating Token Index: 267

Text #18

Max Range: 3.9152. Min Range: -3.9152

Max Act: 1.9853. Min Act: -0.0000

Data Index: 1069595 (C4 (Web Text))

Max Activating Token Index: 941

Text #19

Max Range: 3.9152. Min Range: -3.9152

Max Act: 2.0814. Min Act: -0.0000

Data Index: 1085196 (C4 (Web Text))

Max Activating Token Index: 776